Machine-guided design of cell-type-targeting cis-regulatory elements

Area of Science:

• Genomics
• Molecular Biology
• Bioengineering

Background:

• Cis-regulatory elements (CREs) are crucial for controlling gene expression and defining cell types.
• Natural CREs may not be optimal for therapeutic applications requiring precise tissue specificity.

Purpose of the Study:

• To develop a platform for engineering and validating synthetic CREs with programmed cell-type specificity.
• To create novel regulatory elements for potential use in gene therapy and biotechnology.

Main Methods:

• Utilized deep neural network modeling to predict CRE activity.
• Employed in silico optimization and massively parallel reporter assays (MPRAs) to design and test thousands of synthetic CREs.
• Validated synthetic CREs in vitro using cell lines and in vivo in analogous tissues.

Main Results:

• Synthetic CREs demonstrated enhanced cell-type-specific gene expression compared to natural human CREs in vitro.
• In vivo testing confirmed specific expression in target tissues.
• Engineered CREs showed distinct sequence motifs associated with on-target activity and reduced off-target activity.

Conclusions:

• A generalizable framework was established for engineering synthetic CREs using MPRAs.
• The study provides a method to "write" regulatory code for fit-for-purpose applications.
• Synthetic CREs offer a promising avenue for precise control of gene expression in therapeutic and biotechnological contexts.